From: Florian Frank <flori@...>
Date: 2007-06-06T08:29:35+09:00
Subject: Re: [PATCH] Another small bug fix for Ruby 1.9 and a request for clarification

This is a multi-part message in MIME format.
--------------050902050202040508090609
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit

Hello,

Nobuyoshi Nakada wrote:
>> The second makes it
>> possible to build 1.9. with an installed Ruby 1.8. for bootstrapping.
>>     
>
> For cross-compiling?
That's an interesting question. I had a problem with the new syntax in 
the extmk.rb file, but now I see, that it's supposed to be called from 
miniruby. It's likely, that I had some older 1.8 miniruby laying around. 
I guess, I just got confused. The change might be useful for 
cross-compiling as well.

I usually have a hard time compiling 1.9 compared to 1.8. But I guess 
this is to be expected, because development is so much faster now than 
it used to be before the yarv merge.

I have a attached patch for rdoc-dot. The dot program of the graphviz 
package has changed it's version string output in some of the newer 2.x 
versions. Without the patch rdoc cannot draw the nice module/class 
diagrams, because it cannot figure out the required version. It might be 
useful to apply this patch to Ruby 1.8 as well (if the problem wasn't
already corrected there).

Maybe I should write a different email for this, but anyway. NARUSE Yui sent me an email about a failing JSON test. It seems the semantic for these cases have changed from Ruby 1.8.5 to Ruby 1.9:

RUBY_VERSION # => "1.8.5"
for (foo, bar) in [[1,2], [3,4]]
  [foo, bar] # => [1, 2], [3, 4]
end

for foo, bar in [[1,2], [3,4]]
  [foo, bar] # => [1, 2], [3, 4]
end

RUBY_VERSION # => "1.9.0"
for (foo, bar) in [[1,2], [3,4]]
  [foo, bar] # => [1, nil], [3, nil]
end

for foo, bar in [[1,2], [3,4]]
  [foo, bar] # => [1, 2], [3, 4]
end

I fixed my tests by always using the second form, but this could be a 
Ruby bug. Or is the semantic change intended?

-- 
Florian Frank


--------------050902050202040508090609
Content-Type: text/plain;
 name="rdoc-dot.patch"
Content-Transfer-Encoding: base64
Content-Disposition: inline;
 filename="rdoc-dot.patch"

SW5kZXg6IGxpYi9yZG9jL29wdGlvbnMucmIKPT09PT09PT09PT09PT09PT09PT09PT09PT09
PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PQotLS0gbGliL3Jkb2Mv
b3B0aW9ucy5yYgkocmV2aXNpb24gMTI0NDMpCisrKyBsaWIvcmRvYy9vcHRpb25zLnJiCSh3
b3JraW5nIGNvcHkpCkBAIC01NTAsMTIgKzU1MCwxMiBAQAogICAgIHZlciA9IG5pbAogICAg
IElPLnBvcGVuKCJkb3QgLVYgMj4mMSIpIGRvIHxpb3wKICAgICAgIHZlciA9IGlvLnJlYWQK
LSAgICAgIGlmIHZlciA9fiAvZG90XHMrdmVyc2lvbig/OlxzK2d2aXopP1xzKyhcZCspXC4o
XGQrKS8KKyAgICAgIGlmIHZlciA9fiAvZG90Lio/dmVyc2lvbig/OlxzK2d2aXopP1xzKyhc
ZCspXC4oXGQrKS8KICAgICAgICAgb2sgPSAoJDEudG9faSA+IDEpIHx8ICgkMS50b19pID09
IDEgJiYgJDIudG9faSA+PSA4KQogICAgICAgZW5kCiAgICAgZW5kCiAgICAgdW5sZXNzIG9r
Ci0gICAgICBpZiB2ZXIgPX4gL15kb3QgdmVyc2lvbi8KKyAgICAgIGlmIHZlciA9fiAvXmRv
dC4qP3ZlcnNpb24vCiAgICAgICAgICRzdGRlcnIucHV0cyAiV2FybmluZzogWW91IG1heSBu
ZWVkIGRvdCBWMS44LjYgb3IgbGF0ZXIgdG8gdXNlXG4iLAogICAgICAgICAgICJ0aGUgLS1k
aWFncmFtIG9wdGlvbiBjb3JyZWN0bHkuIFlvdSBoYXZlOlxuXG4gICAiLAogICAgICAgICAg
IHZlciwK
--------------050902050202040508090609--